Yes, as lower an CCT and Illuminance that the less significant CRI. The main things is below of BBL. I do not tried optisolis but i do not seen difference between e21 r80/r9080 both 3000k
Clemence, this is a very comprehensive analysis, and I understand most of your points. I am just wondering if there exist an application that requires the extreme cri at 3000k?
At anything lower than 4000K I too don’t see much difference in colour reproduction EXCEPT for the redness. This means 2000K R70/80 is very yellow while 2000K R9080 is very orange. I don’t appreciate the high 90+ in warm CCT, only the higher wavelengths it brings. Until there’s a R7080 then R9080 is the only way to go to get such warmth from WW.
- Clemence
Late incoming shipment from Nichia arrived today:
This is a special E21A CCT I could not get. But surprisingly, this time Nichia has it. Based on my experience, 4500K in high CRI is not so easy to get from Nichia.
It’s D220, lower than I expected. But I see a pattern here. Brighter bin usually has higher voltage. For example earlier E21A sm403 D240 has M1 voltage bin. Voltage rank goes like this: L1 < L2 < M1
So D220 means its approximately as bright as the top bin D220 219B R9080 and lower than D240 219C R9050. It’s very good considering its domeless. Combined in quadtrix it should output approximately equal to the domed E900 144AM/AR.
- Clemence
And 500 pieces is a good stash too!!
I really like your photos
the E21A and Optisolis seem to have beautiful Red content.
I look forward to seeing some peoples mods with your new quad board and High CRI E21A and Optisolis. They have considerably higher R9 than 219C, which seem yellow to me.
I started to strongly think about a thrower build. I found a single cell (26650) Manta Ray C8.2 host, which is compatible with a 6V/3A boost driver, and is available with OP reflector. That host is originally intended for XHP50/50.2, but with a 6V configured VR21SP4, I think, it should work too (maybe I will need to grind the basement of the reflector to bring the focus closer to the MCPCB - but I hope, that will not be necessary).
Based on the 4×E21A measurements, which Texas_Ace made, the 6V/3A will be the exact sweet spot for this quadtrix.
If this will work for me, that will be super awesome… a 95 CRI ~1750 lumen pocketable thrower :sunglasses:
Ah… I just going to receive the current order from the local post, but I already thinking about the next Clemence order… I became an addict.
As requested by some member to post the CCT comparison tint shot of my Tiara E21A built. No 5000K, 5700K, and 6500K ordered by the time the picture was shot. FYI these Tiaras in the pictures below are all taken.
Sorry for the unexpected shipment delay. I had problem with the silicone thermal potting compound curing. I had to remove and clean all the potted Tiaras and start all over again :weary: . This time I made a special mixer to ensure they mixed well (and hardened completely)
Same picture defocused
- Clemence
Below is the test results of the Tiara Pro E21A. The base test used 6500K D240 bin. I don’t have the time to test all the CCTs one by one. But a fairly accurate calculation can be made using Nichia’s flux binning. Output for each CCT depends on the flux bin, adjust accordingly to get yours.
Actual measurement
Tiara Pro 4x E21A sm653 D240 R9080 [Yajiamei 60deg]
T2: 645 lumens
T1: 339 lumens
M3: 154 lumens
M2: 55 lumens
M1: 22 lumens
FF3: 3,58 lumens
FF2: 0,86 lumen
FF1: 0,03 lumens
Approximate output of other CCTs:
- 2000K D160 will have 160/240 = 66% the output of 6500K D240
- 2500K (D160 + D200)/2 = D180 will have 180/240 = 75% the output of 6500K D240
- 3000K D200 will have 200/240 = 83% the output of 6500K D240
- 3500K (D200 + D220)/2 = D210 will have 210/240 = 87,5% the output of 6500K D240
- 4000K D220 will have 220/240 = 92% the output of 6500K D240
- 4500K D220 will have 220/240 = 92% the output of 6500K D240
- 5000K D220 will have 220/240 = 92% the output of 6500K D240
- 5700K (D220 + D240)/2 = D230 will have 230/240 = 96% the output of 6500K D240
- Clemence
Looks very good. Was it actively cooled?
Cooled in a glass jar filled with water. Start temp was ~25°C, ended at ~27°C
- Clemence
D220 is 220-240 lm. So I think it would be good to use 230, not 220 like you calculate. Same with other bins.
I know but in this case being slightly pessimistic is better because of binning fluctuation between LEDs.
- Clemence
Interesting that the E21A’s characteristic is so different from the XPL and 219C using the same driver. Such a nice flat graph. Do you think the gradual lumen decline with the XPL and 219C is due to temperature regulation from the driver or just the way they behave?
For the very first few seconds it could be due to thermal sag. Even it’s water cooled, the temperature raise in the MCPCB can’t instantly released. Then the battery high current drain ability plays a very important role.
Quadtrix E21A in parallel has extremely low forward voltage, maximizing the flat regulated runtime. E21A and Vapcell unleashes Tiara’s true potential. I’m pretty sure the total efficiency of Tiara’s buck driver is way better than the boost driver in Wizard. I like Tiara more than Wizard in this regard. Even without water cooling the output vs size is already optimum for almost any applications.
Unfortunately I don’t have stock Tiara with XPL to test it with Vapcell. But I’m sure the flat regulated runtime wouldn’t be as long as E21A
- Clemence
I just received 2 modded Utorch S1 Mini from Bali.
Clemence did an amazing job of LED swapping and lens frosting. The light on the right is a Rose Gold Olight S Mini just for reference.
Im in love with the 2000k E21A, and the sw45 is no slouch either, basically canned sunlight 
If anyone has any inclination to ask Clemence to mod a light or provide LEDs, I cant say enough good things about the man, his work, and his LEDs… Im a total Fan!
Thank you Clemence!
How hard would it be to mount the E21As in a multi-emitter lens light like M43/D4S/MF01/ROT66 etc?
Thanks for the comparison. I also have mine on order. I like that the sw45 is pinkish more than magenta. The E21A 2000k seems like it would work very well as an ambient light in the bedroom when I watch TV. :+1:
I LOVE the LEDs, I recommend them highly!
more beamshots in my Utorch post:
I just received the 5 Tiara C1 Pros modded with 4x E21A from Virence.com and I am very happy with Clemence’s work! The packaging of my emitter/mcpcb order was phenomenal too! :+1: 
I love the E21A tint especially the 2500k and 3000k version, which is perfect for bedside use. They are rosier tinted than the 3000k 9050 219B bedside EDC I’ve been using, which I already thought looks really nice. The 4000k is slightly rosier looking than my 9050 219Bs but not as rosy as the Cree 5D tints on the Emisar lights. However, it looks noticeably cooler than the Cree 5D tint. I think the 219B 9080 will be even better looking tint wise if you prefer rosier tint. I took lumen readings at 1sec (turn-on) using the TA Tube calibrated with Maukka calibration lights:
Tiara C1 Pro 4000k – 572 lumens
Tiara C1 Pro 3500k – 539 lumens
Tiara C1 Pro 3000k – 529 lumens
Tiara C1 Pro 2500k – 438 lumens
Tiara C1 Pro 2000k – 407 lumens
Lights were running with fully charged Aspire 18350 1100mah batteries. I doubt it matters what batteries I use though since the amps are pretty low. But the output is quite impressive given how small the Tiara is and how good the tint and CRI looks. Keep in mind these are true calibrated OTF lumens, not manufacturer over inflated lumens.